Presenter

Ari Rasch is a PhD student at the University of Münster, Germany. His research focuses on code generation, optimization, compilation techniques, and programming language design for multi- and many-core systems, e.g., equipped with multi-core CPU, accelerator devices such as many-core Graphics Processing Unit (GPU), and/or mobile CPU/GPU. His main research goals are to: 1) automatically generate portable high-performance compute kernels for data-parallel applications on multi- and many-core architectures, 2) develop automatic performance-tuning methodologies (a.k.a. auto-tuning), and 3) simplify kernel execution on distributed and heterogeneous systems. He focuses on important and complex computations, e.g., in the area of deep learning.